HOW TO SET STANDARD FOR ONE MAN TWO MACHINES OPERATION

[TREE342]

HI, EVERYBODY,

I RUN INTO A SITUATION AND DO NOT KNOW WHAT IS THE BEST WAY TO HANDLE IT.
OUR GRINDING DEPARTMENT HAVE TWO NC GRINDING MACHINES OPERATED BY ONE OPERATOR. DEPENDS ON THE SHOP LOAD AND THE AVAILABLITY OF MATERIAL, SOMETIME THE OPERATOR RUN ONE MACHINE, YET OFTEN HE TAKE CARE BOTH MACHINES AT THE SAME TIME. THE OPERATOR IS WORKING 12 HOURS PER DAY.
AS A MANUFACTURING ENGINEER WE SET STANDARD FOR SETUP AND RUN TIME. WE KNOW HOW TO DEAL WITH ONE MAN ONE MACHINE SITUATION. AND WE KNOW HOW TO DEAL WITH ONE MAN TWO MACHINES SITUATION ONLY IF WE KNOW WE ALWAYS HAVE MATERIAL TO KEEP BOTH MACHINES BUSY.
FROM THE SHOP SCHEDULING STANDPOINT, WE CAN USE UTILIZATION PERCENTAGE TO DEFINE OUR PROCESS POTENTIAL CAPACITY. BUT WHAT IS THE FAIR WAY TO MEASURE THIS OPERATOR'S PERFORMANCE IF THAT IS MEASURED BY STANARD TIME DIVIDED BY ACTUAL TIME TURNED IN.
THANK YOU IN ADVANCE IF ANY ADVICE YOU AND GIVE ME.

 

[BillPSU]

Set your standard for 1 man 1 machine if production cannot guarantee materials for both machines most of the time. Its not your job to measure operator performance that is supervision's job. Your job is to set a time standard giving accurate cost information for your product. Supervision has always used the performance against standard as a club against operators and you as a manufacturing engineer have been put on the spot about standards by operators. In the overall scheme operator performance is not the function of manufacturing engineering. Let supervision find a way to deal with the issue of performance.

Now the flip side. Ask you supervisor which way you should set the standard and follow his lead. Being politically correct maybe the best alternative.

The first portion is a rant against all the wasted time spent adjusting time standards for minor changes in cost. Direct labor should have little to do with the cost of a product. Most product cost usually 50-70% comes from purchased materials. Direct labor impacts the product cost between 5-12%. Even if you cut your direct labor in half the savings in the cost of the product is 2.5-6%. Your overhead is the major operating cost.

 

[griffengm]

Tree342,
The time standard you set is as much a measure of supervision and management as a measure of labor.
My preference would be to set the standard at .5 man per machine. It is then for supervision to determine how to best use the other .5. If the supervisor chooses to not assign other tasks then his shift takes the variance.
If you set the standard at one per machine then the argument can be made that a second operator is needed when the second machine is running.
There is a saying that "Work expands to fill the time allotted." Once this happens you will have a real struggle on your hands correcting it.

Griffy

 

[aamoroso]

For CNC Mills we used to measure the job, regardless of manpower. Job A runs 4.5 hours per part, Job B runs 2.5 hours per part... If one man can run both, great, if not no difference, the parts still take as long. Grinding may be a bit difficult to measure parts in progress, but here is what we used to do. Person A comes in on 1st shift and starts on Tool 10, pulls three parts and finishes on Tool 6. After four complete measured cycles of this we could establish the time needed to complete a part. Then we would break down the amount of time spent per tool (to be fair to the operators). The operator goal was to achieve 93% efficiency for all jobs. Every six months, the time was recalculated (dropping the highest and lowest times). This kept the jobs improving, but kept the times reasonable for new operators. Eventually we knew exactly how much time a job had to run, it didn't matter how many operators were running the jobs, just the amount of time it needed in the machine.

Fill what's empty. Empty what's full. And scratch where it itches.

 

[griffengm]

Tree342,
Juran (everyone knows who him, right?) says that there are several types of jobs out there. Three that I recall offhand are machine determinant, operator determinant, and process determinant. Most in our shop are some combination of these.
In the case of an operation where the operator loads the machine, hits a button to start the cycle and has nothing else to do until the cycle ends when he then unloads it, I count the cycle time plus load/unload time. If the time during the machine cycle is not spent on something required for this part, I do not count it. If he uses it to clean the work area or read a magazine it is not direct labor. If he is required to pack parts or verify each part by customer specs, the time gets counted.
Keep in mind that many conventions followed by a company regarding time standards may be developed by management other than or in conjunction with IE. My tendency is to run very tight and force supervision to justify the use of the resources. I will close now because I'm headed toward a rant about supervisors and management not fulfilling their obligation to do their job.

Griffy

 

[MikeNH]

We do it by reducing the machine cost rate. If you use any type of MRP or other resouce planning tools you dont want to change the actual machine time.

We dont split the rate in 1/2. We charge each machine about 60% of the old rate. We have some that because of the type of operation we adjust.

We also use the same logic for 3, 4, and upto 8 machines per operator.

 

[bill1898]

TREE342,
YOUR STANDARD SHOULD REFLECT THE MOST EFFECTIVE WAY OF PRODUCING THE PARTS. IT WOULD SEEM TO BE LESS COSTLY TO HAVE ONE OPERATOR SERVICING TWO MACHINES THAN TO HAVE AN OPERATOR AT EACH MACHINE, ASSUMING THAT THERE IS NOT ANY SUBSTANTIAL MACHINE IDLE TIME AS A RESULT. EVEN THOUGH I AGREE WITH BillPSU THAT DIRECT LABOR IS A SMALL PART OF PRODUCT COST I DO THINK THAT PAYING TWO PEOPLE TO DO THE JOB THAT CAN BE ACCOMPLISHED BY ONE IS NOT GOOD BUSINESS.
THE ABSOLUTE MEASURE OF TOTAL EFFECTIVENESS IS HOW MANY ACCEPTABLE PARTS ARE PRODUCED WITH MINIMAL INPUT. YOUR TASK IS TO SET A REASONABLE STANDARD BASED UPON THE MACHINE AND OPERATOR CAPABILITIES. THE ACTUAL OUTPUT ACHIEVED IS THE RESPONSIBILITY OF THE MANAGEMENT PEOPLE INVOLVED AS PREVIOUSLY NOTED.

 

[sacem1]

You must draw bar diagrams reflecting the load/operate/unload work sequence of each machine, then making a timetable of repeated operations you will find how much time the worker part (load/unload) superimposses on both machines, then you can analize this:

1)If it never does you can divide the worker in two.

2) If you do have simultaneous tasks you must determine how much time does any machine get unnecesarily delayed if you have = slack (unused time in the machine workload) then that one gets pryority assigment of the worker, but if both have zero slack then you have to determine either by cost or productivity which machine you delay.

3) Once your delayed times have been stablished you have to compare cost/benefit of stopping one machine or hiring another worker

4)If machine time slack was present then you have zero total delay and you can go on with one worker and act as in situation 1

And that's it

Cheers

SACEM1